Use, misuse, and ambiguity of indices of residence in acoustic telemetry studies

Habitat associations and preferences of animals can be inferred from how long they remain within close proximity to a certain location. The residency index (RI) is a common metric used in acoustic telemetry studies to assess how long an individual spends in an area. However, the methods used to calculate RI can affect the interpretation of telemetry results. The index has been used under different names and with different equations and definitions. This review of 72 publications highlights that RI has been defined using 2 main equations: RI_A, which divides the number of days detected by days at liberty; and RI_B, which divides the number of days detected by the monitoring period. We present a case study using long-term acoustic telemetry data from 244 individuals of 8 species collected in Queensland, Australia, to assess how the definition of RI affects ecological interpretation. Over 3 million detections from Carcharhinus amblyrhynchos, C. melanopterus, Galeocerdo cuvier, Hemigaleus australiensis, Lethrinus miniatus, L. nebulosus, Lutjanus sebae, and Plectropomus leopardus were analysed to evaluate how the 2 main RI equations differ in the results they provide and their possible interpretation. The 2 equations yielded significantly different RIs for some species and individuals. This was primarily driven by variation in behavioural ecology. Either equation can be useful depending on the aim of the study. However, we propose using both equations in combination to better identify detection patterns and strongly recommend future studies clearly define the equation to enable comparisons and appropriate interpretation of results.